Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language...

21
Title <Notes>Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb-- Author(s) Vattukumpu, Tero Citation 京都大学言語学研究=Kyoto University Linguistic Research (2018), 37: 41-60 Issue Date 2018-12-31 URL https://doi.org/10.14989/240978 Right © 京都大学言語学研究室 Type Departmental Bulletin Paper Textversion publisher Kyoto University

Transcript of Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language...

Page 1: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

Title <Notes>Notes on reduplication in the Haisla language --Partialreduplication of the negative auxiliary verb--

Author(s) Vattukumpu, Tero

Citation 京都大学言語学研究=Kyoto University Linguistic Research(2018), 37: 41-60

Issue Date 2018-12-31

URL https://doi.org/10.14989/240978

Right © 京都大学言語学研究室

Type Departmental Bulletin Paper

Textversion publisher

Kyoto University

Page 2: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

Notes on reduplication in the Haisla language Partial reduplication of the negative auxiliary verb

Tero Vattukumpu

Abstract: In this paper I will point out that, according to my data, a plural form of the negative auxiliary verb formed by means of partial reduplication exists in the Haisla language in contrary to a description on the topic in a previous study. The plural number of the subject can, and in many cases must, be indicated by using a plural form of at least one of the components of the predicate, i.e. either an auxiliary verb (if there is one), the semantic head of the predicate or both. Therefore, I have examined which combinations of the singular and plural forms of the negative auxiliary verb and different semantic heads of the predicate are judged to be acceptable by native speakers of Haisla. My data suggests that – at least at this point – it seems to be impossible to make convincing generalizations about any patterns according to which the acceptability of different combinations could be determined *.

Keywords: Haisla, partial reduplication, negative auxiliary verb, plural, root extension

1 Introduction

The purpose of this paper is to introduce some preliminary notes about reduplication in the

Haisla language from a morphosyntactic point of view as a first step towards a better

understanding and a more comprehensive analysis of all the possible patterns of reduplication

in the language. The focus of this paper is on the following two points: (a) The plural form of

the negative auxiliary verb formed by means of partial reduplication, the existence of I will

point out in contrary to some accounts of the topic made in the previous descriptions of the

language (Bach, Robinson & Robinson n.d., Lincoln & Rath 1986a: 233) (b) The

co-occurrence of the plural and singular forms of the negative auxiliary verb and the semantic

head of the predicate in intransitive declarative clauses.

* All the data used in this paper is my own data collected during two field trips: September 1 –

October 31, 2017 and June 27 – August 31, 2018, unless indicated otherwise. Both trips were made financially possible by the Nomura Foundation which has funded my studies since 2017. All the data were kindly provided by Mr. Nelson Grant and another consultant who wishes to remain anonymous. I would like to thank both of my consultants for their valuable teachings and willingness to help me with my research as well as two anonymous reviewers for their constructive comments.

京都大学言語学研究 (Kyoto University Linguistic Research) 37 (2018), 41 –60

- 41-

Page 3: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

In this chapter I will briefly provide basic information about the Haisla language including

some genealogical and geographical information, as well as a typological profile and some

basic linguistical information of the language from mainly a morphological point of view.

After this I will introduce the topic of this paper in more detail.

In the second chapter I will introduce a study done previously on reduplication in Haisla by

Bach (1990a) and point out the role that reduplication plays in expressing plurality in Haisla.

After that the rest of the paper will deal with the plural form of the negative auxiliary verb

formed by partial reduplication, actual examples and data, and the author’s view on the

possible interpretations of it.

Haisla is one of the several indigenous languages of Canada and is spoken by the Haisla

first nation who resides on the Pacific coast in the central part of the province of British

Columbia. The language is nowadays spoken primarily in the village of Kitamaat as well as in

the surrounding area and by some members of the community who live in a diaspora

elsewhere. Haisla belongs to the Northern Wakashan group of languages of the Wakashan

language family. Haisla is a severely endangered language with 87 fluent speakers left1.

Haisla has a basic Predicate-Subject-Object word order with a strong tendency of being

head-marking. It is morphologically a very complex and almost exclusively agglutinating

language with an extremely large inventory of derivational suffixes2 out of which quite many

are so called lexical suffixes that carry a rather lexical meaning. Every word consists of a root

(morpheme) which usually carries at least one suffix with it (Vink 1977: 122). Since long

strings of suffixes attached to one root are possible, there are words with relatively high

morphemes-per-word ratio. In addition to this, Haisla also has a polypersonal marking on its

predicates3 and an elaborated deixis system4 which is also manifested by obligatory deixis

1 Information obtained by personal communication with Mrs. Teresa Windsor, Haisla Nation Council’s

Community Cultural Coordinator (2018). 2 According to Bach (1995: 13), there are around 500 derivational suffixes in Haisla. 3 To be exact, there might be room for dispute over whether the personal markers should always be

analyzed as being placed within the predicate because they are always attached to any clause-initial item, usually an auxiliary verb (AUX), unless there are both subject and object personal markers present in which case the object marker attaches to a word following the clause-initial item. These clause-initial items can sometimes be analyzed as an item outside the predicate, e.g. conjunctives. Because of this kind of morphosyntactic behavior of the person markers and the fact that they are absent whenever there is either a personal pronoun or a noun phrase in the subject position of the clause, I have adopted the view that they are (anaphorical) (en)clitics instead of person agreement suffixes.

4 The deixis system of Haisla has a four-way distinction with the following four deictic categories: proximal, medial, distal and a forth one which I call absental. The first three categories are the conventional ones which can be found in many other languages that have a deixis system with three deictic categories, i.e. ‘near the speaker’ (proximal), ‘near the listener’ (medial) and ‘not near to either one of them’ (distal). Bach (2006: 267, 279) describes the fourth deictic category as a “spatiotemporal”

Tero VATTUKUMPU

- 42-

Page 4: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

suffixes, and these have together led to a sizable set of inflectional suffixes as well. Hence, it is

not surprising that Haisla has been described as a polysynthetic language (Bach 1995: 13).

However, the complexity of Haisla morphology does not rely solely on heavy suffixation as

there can also be observed pervasive modification of the roots, which I call root extensions5

and which manifests itself mainly in the form of partial reduplication. Some of the patterns of

reduplication on Haisla roots have been described in detail in previous studies, one of which

by Bach (1990a) will be introduced in the next chapter.

Due to the omnipredicative nature of the Haisla language, a root from any lexical category

can, in general, be used to form a stem6 to be used as the predicate of a clause. For example,

stems semantically corresponding to English nouns can act as a predicate with the meaning ‘to

be X’. These stems that carry most of the lexical contents of the predicate are referred to as the

semantic head (of the predicate) in this paper. A predicate may consist of only the semantic

head or it can also be preceded by one or more of the many auxiliary verbs (which also

include the negative auxiliary verb). In this paper four criteria – both morphological and

syntactic – are used to define an auxiliary verb in the Haisla language. An auxiliary verb is a

word that fulfills all the following criteria:

(I) It can form a predicate with a semantically noun- or verb-like semantic head of the

predicate which carries most of the lexical meaning of the predicate.

(II) It precedes the semantic head of the predicate.

(III) It takes the anaphorical subject enclitic instead of the semantic head of the predicate

unless it follows another auxiliary verb in which case the preceding auxiliary verb

takes the enclitic.

(IV) The only allomorph of the anaphorical clitic for the exclusive 1st person singular that

it can take is =n.

Examples of predicates with and without an auxiliary verb can be seen below.

on the grounds that it does not seem to refer only to spatial deixis, unlike the other three categories, but also to a temporal one having a meaning “just gone”. According to Bach (2006: 279) and my own field data, the speakers often translate clauses with absental referent into an English clause that is in a past tense. Bach (2006: 268, 273, 279-280) also points out that there is an additional distinction between visible and invisible in all the deictic categories except for absental. Since I have not managed to collect data that could confirm the existence of the distinction based on visibility, I have excluded it from the deixis system in this paper.

5 Also called “stem extensions” by Bach (1990a). 6 As far as I know, a stem can consist of either one root, or a root with one or more derivational

suffixes.

Notes on reduplication in the Haisla language ̶ Partial reduplication of the negative auxiliary verb ̶

- 43-

Page 5: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

(1) gélw̓a gáda (2) w̓ab-ád=nugʷa

canoe DEM.PROX water-LS.to.have=1EXCL.SG

‘This is a canoe.’ ‘I have water.’

(3) bibíʼu-∅-enc gída (4) ʼíkuy=i qáɬela

uncle-DEIX.DIS-POSS.1EXCL.SG DEM.DIS AUX.CAP=3SG/PL.MED walk

‘That (over there) is my uncle.’ ‘(S)he (near you) can walk.’

Plurality of the subject of a clause can be encoded by using a plural form of at least one of

the components of the predicate, i.e. either an auxiliary verb, the semantic head of the

predicate or both as seen in the examples below.

(5) ʼiʼikuya=nis7 qáɬela (6) ʼíkuya=nis qiqáɬela

AUX.CAP.PL=1INCL.PL walk AUX.CAP=1INCL.PL walk.PL

‘We (INCL) can walk.’ ‘We (INCL) can walk.’

(7) ʼiʼikuya=nis8 qiqáɬela

AUX.CAP.PL=1INCL.PL walk.PL

‘We (INCL) can walk.’

In the later chapters, after showing that the plural form of the negative auxiliary verb

formed by means of partial reduplication exists in the first place, I will show what kind of

combinations of the plural and singular forms of the negative auxiliary verb and the semantic

head of the predicate are possible, and which of the combinations are allowed as acceptable

expressions in the case of different semantic heads according to my data. After that, I will

examine whether any generalizations can be made to describe patterns according to which all

the acceptable combinations could be predicted.

The orthography used in this paper is a slightly modified version of the orthography used by

Bach (2001: 72). All the consonant phonemes and graphemes of the orthography used in this

paper are listed below9.

7 The place of stress is not confirmed. 8 The place of stress is not confirmed. 9 The consonant phonemes with the greatest allophonic variation are /p/, /t/, /kʲ/, /kʷ/, /q/, /qʷ/, /t͡ s/ and

/tˡ/ which have the following allophones: [p~b], [t~d], [kʲ~ɡʲ], [kʷ~ɡʷ], [q~ɢ], [qʷ~ɢʷ], [t͡ s~d͡z] and [tˡ~dˡ].

Tero VATTUKUMPU

- 44-

Page 6: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

/p/: ⟨b⟩ /t͡ s/: ⟨z⟩ /s/: ⟨s⟩ /kʷ/: ⟨gʷ⟩ /qʼ/: ⟨q̓⟩ /ʔ/: ⟨ʼ⟩ /l/: ⟨l⟩ /pʰ/: ⟨p⟩ /t͡ sʰ/: ⟨c⟩ /ɬ/: ⟨ɬ⟩ /kʷʰ/: ⟨kʷ⟩ /χ/ː ⟨x̄⟩ /h/: ⟨h⟩ /lʼ/: ⟨l̓⟩ /pʼ/: ⟨p̓⟩ /t͡ sʼ/: ⟨c̓⟩ /kʲ/: ⟨g⟩ /kʷʼ/: ⟨k̓ʷ⟩ /qʷ/: ⟨ḡʷ⟩ /m/: ⟨m⟩ /j/: ⟨y⟩ /t/: ⟨d⟩ /tˡ/: ⟨λ⟩ /kʲʰ/: ⟨k⟩ /xʷ/: ⟨xʷ⟩ /qʷʰ/: ⟨qʷ⟩ /mʼ/: ⟨m̓⟩ /jʼ/: ⟨y̓⟩ /tʰ/: ⟨t⟩ /t͡ ɬ/: ⟨ƛ⟩ /kʲʼ/: ⟨k̓⟩ /q/: ⟨ḡ⟩ /qʷʼ/: ⟨q̓ʷ⟩ /n/: ⟨n⟩ /w/: ⟨w⟩ /tʼ/: ⟨t̓⟩ /t͡ ɬʼ/: ⟨ƛ̓⟩ /xʲ/: ⟨x⟩ /qʰ/: ⟨q⟩ /χʷ/: ⟨x̄ʷ⟩ /nʼ/: ⟨n̓⟩ /wʼ/: ⟨w̓⟩

The most recent and common views of the vowel phonemes of Haisla holds that there are

three phonemic vowels (Lincoln & Rath 1986a: 1-34, Bach 2001: 72). I have adopted this

view in this paper. The vowel phonemes and the graphemes used to write them in this paper

are the following10.

/a/: ⟨a⟩ /i/: ⟨i⟩ /u/: ⟨u⟩

Epenthetic vowels are used extensively to break consonant clusters, but these phonetic

vowels have not been treated as phonemes in the major analyses made previously on the

phonological system of the language since their occurrence is mostly predictable (Vink 1977,

Lincoln & Rath 1986a). I have adopted this view in this paper. Various epenthetic vowels –

referred collectively to as schwa in this paper – are written as ⟨e⟩. The stress – the location of

which can be either a vowel or a syllabic resonant marked by a combination of a schwa and a

resonant grapheme in the orthography – is marked with an acute accent.

2 Reduplication in Haisla

In this chapter I will introduce the analysis of Bach’s (1990a) work on the topic of

reduplication in Haisla. This is the main work done previously on this topic. It is intended to 10 All the three vowel phonemes have allophones, some of which show only a slight difference from

the most basic phonetic value of the phoneme, but coalescence of vowel phonemes in a same sequence and diphthongization also take place in certain environments (Lincoln & Rath 1986a: 1-34). Two of the most important of such changes are the diphthongization of the high vowels (/i/ and /u/) after any uvular or glottal consonant, and the coalescence of /ai/ into [ɛ(ː)] and /au/ into [ɔ(ː)]. After uvular and glottal consonants, and to some extent also some ejectives of a different place of articulation, /i/ and /u/ become closing diphthongs resembling [ei̯] and [ou̯] respectively, with slight variation in the actual phonetic value (Vink 1977: 113). Sequences of the same vowel phoneme are pronounced in a different way depending on the phoneme (Lincoln & Rath 1986a: 22-24). Only /aa/ becomes a long vowel ([aː]) creating a length opposition between /a/ and /aa/, depending on whether stressed or not and on the phonetic environment (Lincoln & Rath 1986a: 24). When two different vowel phonemes form a vowel sequence there is usually a phonetic glide ([w] or [j]) pronounced between them (Lincoln & Rath 1986a: 1-34).

Notes on reduplication in the Haisla language ̶ Partial reduplication of the negative auxiliary verb ̶

- 45-

Page 7: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

describe the main patterns of root extensions of the Haisla roots recognized so far which I will

summarize in this chapter. I will also give a brief overview of how partially reduplicated roots

used as (a part of) the semantic head of the predicate functions to express plurality depending

on the person of the subject.

Bach (1990a: 2-3) lists the following nine types (Table 1) including one type with two

subtypes as “the main types of stem extensions that occur in Haisla”. The examples of the

expanded roots given by Bach (1990a: 2-3) include stems which have been formed by

attaching (derivational) suffixes to the expanded root. These suffixes are not dealt in detail

here as they are not relevant since the main point is the extension of the roots.

Table 1 – The main types of root extensions according to Bach (1990a)

Type Root extension Example

Type 1 Full reduplication of the whole root. bik- ‘transfer fire’11

> bíxbika12 ‘lightning’

Type 2 A reduced form of the root (two consonants with

a schwa between them) following the full root.

cix̄ʷ-13 ‘crabapple’ / ‘sour’

> cíx̄ʷcex̄ʷa ‘eat crabapples’

Type 3 The first consonant of the root repeated and

followed by a schwa.

c̓en- ‘move in mass’/’tidal rapids’

> c̓ec̓én ‘rapids’

Type 4a

The first consonant of the root repeated and

followed by /i/ with the stress in its original

position.

p̓al- ‘work’/’lay hands on’

> p̓ip̓ála ‘work (PL)’

Type 4b

The first consonant of the root repeated and

followed by /i/ with the stress shifting to the new

(first) syllable.

bekʷ- ‘human’

> bíbegʷanem14 ‘people’

11 Bach (1990a: 2) seems not to be sure about the given meaning of the root. However, Lincoln & Rath

(1986b: 429) give the root the same meaning ‘to transfer fire’. 12 The final consonant of the first component of the reduplicated stem has changed (k → x) due to a

regular sound alternation which dictates that aspirated velar and uvular consonants in a morpheme-final position change to homorganic fricatives when they are followed by a consonant or situated in a word-final position.

13 Bach (1990a: 2) gives the root cixʷ- which ends in a labialized velar fricative xʷ, but this should be viewed as a mistake as he uses its uvular counterpart x̄ʷ in the example word. Also, according to Lincoln & Rath (1986a: 430), the root-final consonant should be a labialized uvular fricative.

14 The root-final kʷ has changed to gʷ due to a sound alternation called end effect. The end effect is a morphophonological phenomenon which means that certain suffixes cause certain preceding consonants to undergo a certain change (Bach 2001: 57-58). Aspirated stops like kʷ are always targets of the end effect, and in the example of Type 4b the root-final kʷ has lost its aspiration due to an end effect.

Tero VATTUKUMPU

- 46-

Page 8: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

Type 515

The first consonant of the root repeated and

followed by /a/ with the stress shifting to the

new (first) syllable.

bel- ‘prevent’

> bábel̓a16 ‘sexually jealous’

Type 6

The initial consonant and the vowel of the root

followed by a reduced form of the root (the two

consonants of the root separated by a schwa).

duqʷ- ‘see’/’look’

> dúdeqʷela ‘see (PL)’

Type 7

Modification of the internal shape of the root:

The nature of the modification depends on the

shape of the root.

c̓ip-17

> c̓áipiu18 ‘cedar waxwing’

Type 8 Repetition of the first consonant with two /a/

vowels inserted.

ḡek- ‘woman’

> ḡáḡak̓a19 ‘try to get a wife’

Type 9

One of the other types of reduplication (listed

above) together with the insertion of a new

consonant that can be s, ɬ, x or ʼ.20

buqʷ- ‘bud’ / ‘knob’

> búsbeqʷa ‘sea anemone’

Bach (1990a: 3) also mentions that the types of reduplication listed above in Table 1

sometimes occur with alternations in the first consonants21 and that there are other types of

root extensions too in addition to the nine listed above, for example a type in which a vowel is

inserted into a root-final consonant cluster.

The forms resulting from one of the above-mentioned root expansion types involving

reduplication are used for many functions like expressing plurality or iterative aspect (Bach

1990a: 3). The function of the reduplication in Haisla dealt with in this paper is the first one,

15 Bach (1990a: 2) gives the following description of Type 5: “…just like Type 4a, but the new vowel is

/a/ instead of /i/.” This would mean that the stress does not shift to the first syllable. However, the example Bach (1990a: 2) gives (bábel̓a) has its stress in the first syllable. Lincoln & Rath (1986a: 53) also describes the same lexeme (bábel̓a) as having its stress in the first syllable. Judged from this example, it would seem to be more correct to say that Type 5 should be described to be just like 4b (instead of 4a), unless there were some reason for describing Type 5 as being like Type 4a not mentioned by Bach (1990a), which is why I made the description as it is in Table 1.

16 The root-final l has been glottalized due to an end effect (cf. footnote 14). 17 Bach (1990a: 3) does not indicate the meaning of the root in this example, but according to Lincoln

& Rath (1986b: 433) the root’s meaning is ‘wide’. 18 Bach (1990a: 3) also gives the following examples (without the meanings of the roots) of this type: ḡux̄ʷ- > ḡáux̄ʷeksala ‘squander’ / ‘scatter berries’ c̓em- > c̓ámaksala ‘point aimlessly’ k̓ix̄ʷ- > k̓áux̄ʷeksala ‘run in all directions’ das- > dáaseksala ‘play around diving’ 19 The root-final k has been glottalized due to an end effect (cf. footnote 14). 20 Bach (1990aː 3) notes that the Type 9 should be further subdivided according to the inserted

consonant and the basic type of expansion of the root involved and that all these subtypes have been lumped into Type 9 as they are rather rare patterns and not very productive.

21 Common alternations being s y , ƛ l and ɬ l (Bach 1990a: 3).

Notes on reduplication in the Haisla language ̶ Partial reduplication of the negative auxiliary verb ̶

- 47-

Page 9: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

i.e. plurality. In the next section we will have a look at how forms of roots formed by means of

(partial) reduplication are used to express plurality of the subject in intransitive declarative

clauses.

2.1 Reduplication and plurality

First, let’s have a look at the inflectional system of the subject enclitics in intransitive

declarative clauses. When there is no subject pronoun or noun phrase in a clause, an

anaphorical enclitic indicating the person of the subject is attached to the predicate. The clitics

are listed below in Table 2.

Table 2 – The subject enclitics of the Haisla language

SINGULAR PLURAL

1st person – Exclusive =n / =nugʷa22 =nuxʷ

1st person – Inclusive =nis

2nd person =su =su

3rd person23 – Proximal =ix =ix

3rd person – Medial =u =u

3rd person – Distal =i =i

3rd person – Absental =gi =gi

As can be seen from Table 2, singular and plural share the same enclitic in 2nd and 3rd

persons. However, as mentioned in the first chapter, plurality of the subject can be expressed

by changing the predicate into a plural form, i.e. by using a plural form of either the semantic

head of the predicate, the auxiliary verb, or both. There are stems that do not have a distinct

plural form, but for those that have, there are two methods to form it. One method is to use a

suppletive plural form and the another one is to form a plural form by means of a root 22 The anaphorical enclitic for the exclusive 1st person singular has two allomorphs =n and =nugʷa

which are in free variation excluding any occasion when the enclitic is added to an auxiliary verb as pointed out in the previous chapter.

23 The deixis system of Haisla manifests itself inflectionally in the 3rd person resulting into four subcategories of the 3rd person, reflecting the four deictic categories of Haisla: proximal, medial, distal and absental. However, the system is clearly going through a process leading to a simplification of it as some of the anaphorical subject and object clitics are often used to indicate more than one of the deictic categories. Especially the clitics indicating the medial category seems to be taking over the proximal and distal categories according to my own data. Even though all the anaphorical clitics and forms in the paradigm of the negative auxiliary verb listed in this paper have been labeled according to each of the four deictic categories, it must be acknowledged that some of them can be used also for other deictic categories too.

Tero VATTUKUMPU

- 48-

Page 10: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

extension. This is where (partial) reduplication comes into the picture as it is by far the most

common type of root extension as seen in the previous section, which is why many of the

plural forms that exist there are formed by means of (partial) reduplication.

Because the 1st person and the other two persons differ in whether they have separate

enclitics for singular and plural, the only way to express plurality of the subject in 2nd and 3rd

persons is to use the plural form in the predicate. If there is no partial reduplication on the

negative auxiliary verb while the plural form of the following semantic head is not used in the

case of 2nd and 3rd persons, the default interpretation of the number of the subject would be

singular24.

When the semantic head of the intransitive predicate is accompanied by an auxiliary verb,

there are potentially three different possibilities of expressing plurality in the 2nd and 3rd

persons since there are three possible combinations of singular and plural forms of the two if

at least one of them must be in the plural form. The number of potential combinations in the 1st

person is four since it is possible to use the singular form of both the auxiliary verb and the

semantic head of the predicate because there is no need to use any plural forms to express the

plurality in the 1st person.

In the next chapter I will not only show that there exists a plural form of the negative

auxiliary verb formed by means of partial reduplication, but also examine which combinations

of the singular and plural forms of the negative auxiliary verb and the semantic head of the

predicate are judged to be acceptable by the native speakers.

3 The partial reduplication of the negative auxiliary verb

One of the items fulfilling the criteria for an auxiliary verb is the negative auxiliary verb

which is discussed in this chapter. Bach, Robinson & Robinson (n.d.) seem to list two

allomorphs k̓u(u)- and k̓u(u)s- for the root of the negative auxiliary verb. Lincoln & Rath

(1986b: 461) also give the root k̓uu-25. Even though there might be some room for dispute over

the root form of the negative auxiliary verb, this topic is not discussed in this paper because –

as seen in this chapter – it does not have an effect on the type of root expansion (i.e. partial

reduplication) on the negative auxiliary verb discussed in this paper. The full paradigm of the

negative auxiliary verb according to Bach, Robinson & Robinson (n.d.) and Bach (1990b: 82)

is shown below. 24 Excluding all the instances when the semantic head is (derived from) a root that has no distinct plural

form. 25 Listed as “k̓ww-” according to the phonemic script used by Lincoln & Rath (1986a, 1986b).

Notes on reduplication in the Haisla language ̶ Partial reduplication of the negative auxiliary verb ̶

- 49-

Page 11: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

Table 3 – The paradigm of the negative auxiliary verb according to Bach, Robinson &

Robinson (n.d.) and Bach (1990b)

SINGULAR PLURAL

1st person – Exclusive k̓u(u)=n k̓ú=nuxʷ

1st person – Inclusive k̓ú=nis

2nd person k̓ús=su26 k̓ús=su

3rd person – Proximal k̓ús=ix k̓ús=ix

3rd person – Medial k̓ús=u k̓ús=u

3rd person – Distal k̓ús=i k̓ús=i

3rd person – Absental k̓ús=gi k̓ús=gi

Some examples of the forms in Table 3 are shown below.

(8) k̓u=n m̓ál̓inix̄ʷ (9) k̓ús=i m̓ál̓inix̄ʷ

AUX.NEG=1EXCL.SG hunter AUX.NEG=3SG/PL.DIS hunter

‘I am not a hunter.’ ‘(S)he (over there) is not a hunter.’

(10) k̓ú=nuxʷ máɬela (11) k̓ús=su máɬela

AUX.NEG=1EXCL.PL swim AUX.NEG=2SG/PL swim

‘We (EXCL) are not swimming.’ ‘You (SG) are not swimming.’

There does not exist any description of possible root expansions of the negative auxiliary

verb root known to the author of this paper, but there is a note by Bach, Robinson & Robinson

(n.d.) stating that the plurality of the subject in intransitive declarative clauses is expressed by

using the plural form of the semantic head – whenever there is a distinct plural form available

– in the case of a negative predicate and there is no mention of a plural form of the negative

auxiliary verb formed by partial reduplication or any other means. There is also no mention of

any partially reduplicated form in a detailed description of the negative auxiliary verb by

Lincoln & Rath (1986a: 233).

However, I have managed to find examples of a partially reduplicated form of the negative

auxiliary verb – which I consider to be a plural form of the negative auxiliary verb – when

collecting paradigmatic data of declarative intransitive sentences with negative predicates by 26 The pronunciation is k̓úcu as k̓us- + =su triggers the following sound alternation: s + s → c (Bach &

Bates 1971: 8).

Tero VATTUKUMPU

- 50-

Page 12: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

elicitation. These forms occurred when the referential subject of the sentence was plural.

Examples of the plural form of the negative auxiliary verb formed by partial reduplication

with the subject enclitics of all the grammatical persons have been found. The pattern of the

reduplication in each of these examples is that of either Type 4a or 4b introduced in the

previous chapter as the first consonant of the root is repeated with the vowel /i/. The exact type

still needs to be determined as I still have not managed to create a consistent analysis of the

stress of all the forms found. Therefore, I have not indicated the stress for each of the

combinations of the plural form of the negative auxiliary verb and a subject enclitic that I have

found shown below.

1st person plural – Exclusive k̓ik̓u=nuxʷ 3rd person plural – Proximal k̓ik̓us=ix

1st person plural – Inclusive k̓ik̓u=nis 3rd person plural – Medial k̓ik̓us=u

2nd person plural k̓ik̓us=su27 3rd person plural – Distal k̓ik̓us=i

3rd person plural – Absental k̓ik̓us=gi

Some examples of these forms are shown below.

(12) k̓ik̓us=i hená (13) k̓ik̓u=nis gélt̓ex̄d

AUX.NEG.PL=3SG/PL.DIS sing AUX.NEG.PL=1INCL.PL tall

‘They are not singing.’ ‘We (INCL) are not tall.’

(14) k̓ik̓u=nuxʷ máɬela (15) k̓ik̓us=su máɬela

AUX.NEG.PL=1EXCL.PL swim AUX.NEG.PL=2SG/PL swim

‘We (EXCL) are not swimimg.’ ‘You (PL) are not swimimg.’

The most interesting findings about these forms is not only the fact that they exist and are

used spontaneously, but also the observations about the accepted co-occurrences with the

singular and plural forms of the semantic heads of the predicate depending on the semantic

head used together with them. This will be discussed in the next section.

According to my data, the actual possible combinations of the singular and plural forms of

the negative auxiliary verb and the semantic head of the predicate are the following.

27 The pronunciation is k̓ik̓ucu being a result of the same sound alternation that causes the form k̓úcu,

i.e. s + s → c.

Notes on reduplication in the Haisla language ̶ Partial reduplication of the negative auxiliary verb ̶

- 51-

Page 13: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

Table 4 – The possible combinations of the negative auxiliary verb and the semantic

head of the predicate with a plural subject

Person of the plural subject Possible combinations

The negative auxiliary verb The semantic head

1st person – Exclusive (SG) k̓ú=nuxʷ SG: PL:

(PL) k̓ik̓u=nuxʷ SG: PL:

1st person – Inclusive (SG) k̓ú=nis SG: PL:

(PL) k̓ik̓u=nis SG: PL:

2nd person (SG) k̓ús=su SG: PL:

(PL) k̓ik̓us=su SG: PL:

3rd person – Proximal (SG) k̓ús=ix SG: PL:

(PL) k̓ik̓us=ix SG: PL:

3rd person – Medial (SG) k̓ús=u SG: PL:

(PL) k̓ik̓us=u SG: PL:

3rd person – Distal (SG) k̓ús=i SG: PL:

(PL) k̓ik̓us=i SG: PL:

3rd person – Absental (SG) k̓ús=gi SG: PL:

(PL) k̓ik̓us=gi SG: PL:

This list of the actual possible combinations above matches perfectly – but unsurprisingly –

with the notion made in the previous chapter that when it comes to the potential combinations

of the plural and singular forms of the negative auxiliary verb and the semantic head of the

predicate, there are four combinations in the 1st person and three combinations in the 2nd and

3rd persons. Next, let’s have a look at which of the actual possible combinations listed in Table

4 are judged to be acceptable with certain semantic heads of the predicate.

3.1 Acceptable combinations of the negative auxiliary verb and the semantic head

It can be observed from my data, that the number of the combinations that are actually

allowed to be used as acceptable expressions, depends greatly on the semantic head of the

predicate as well as the consultant. The number of stems whose (almost) full paradigm of the

occurrence with the negative auxiliary verb that were possible to collect as reliable data was

not large. These ten stems are listed below28.

28 The stress is not indicated when its position is unclear.

Tero VATTUKUMPU

- 52-

Page 14: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

m̓ál̓inix̄ʷ ‘hunter’ him̓ás ‘chief’ gélt̓ex̄d ‘tall’

c̓ek̓ʷex̄d ‘short’ kʷíxq̓a ‘angry’ y̓ák̓iqela ‘sad’

hená ‘to sing’ t̓epáƛ29 ‘to fish’ (FUT) máɬela ‘to swim’

dálanuxʷ ‘to have money’

To get straight to the point, it is probably reasonable to think that it is very difficult to make

any generalizations of the patterns of the acceptable combinations of the singular and plural

forms of the negative auxiliary verb and the semantic head of the predicate that each of the

above-mentioned stems allow to be used, not only because the number of stems whose full

paradigms have been possible to collect is rather small, but also because of the following two

reasons: (I) There clearly is a certain amount of uncertainty about how sure the consultants are

about the acceptability of each combination. (II) The semantic variation of the stems is still

perhaps too poor, even though there are stems denoting action/movement (verb-like stems),

quality (adjective-like stems) and entity (noun-like stems). (III) The variation of the patterns of

acceptable combinations between different stems is too varied to make strong generalizations,

as seen below.

I have listed all the acceptable combinations of the singular and plural forms of the negative

auxiliary verb and the semantic head of the predicate for each of the ten stems listed above

(Table 5, see below)30. The corresponding plural form of each stem is indicated below the

singular form of the stem. All the stems have been divided into three semantic categories

according to their meaning: verb-like stem (V), adjective-like stem (A) and noun-like stem (N).

Also, two sets of combinations for one stem have been shown separately when a paradigm has

been obtained from both consultants. All the acceptable combinations have been listed

separately for each person with a formula in which SG stands for the singular form of the

negative auxiliary verb or the semantic head of the predicate and PL for the plural form. The

subcategories of the 1st and 3rd persons have not been shown separately unless the

subcategories have different sets of combinations. When there is a need to differentiate the

subcategories, the following letters have been used: E (exclusive), I (inclusive), P (proximal), M

(medial), D (distal) and A (absental). An SG or PL placed on the left side of the plus sign (+) in

the formula indicates the form of the negative auxiliary verb while one on the right side

indicates the form of the semantic head of the predicate. The combinations have been

29 The stem t̓epáƛ is carrying a future tense suffix -ƛ. 30 All data presented on Table 5 has been obtained by elicitation by asking the consultants to evaluate

the acceptability of each combination.

Notes on reduplication in the Haisla language ̶ Partial reduplication of the negative auxiliary verb ̶

- 53-

Page 15: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

presented with only one formula per person whenever there is data of the combinations with

the singular or plural form of the negative auxiliary verb only, but when there is data of the

combinations with both singular and plural forms of the negative auxiliary verb there are two

formulas separated with a vertical bar ( ) so that all the combinations with the singular form

of the negative auxiliary verb are placed on the left side of the vertical bar and all the

combinations with the plural form of the negative auxiliary verb on the right side31. Any

details are explained in the footnotes.

Table 5 – All the acceptable combinations of the ten stems

Stem Semantic

category

Acceptable combinations

Consultant A Consultant B

m̓ál̓inix̄ʷ

‘hunter’

PL: m̓im̓ál̓inix̄ʷ

N

1PL: SG + PL PL + PL

2PL: PL + PL

3PL: PL + PL32

him̓ás

‘chief’

PL: hiʼemás33

N

1PL: SG + SG/PL

PL + SG/PL34

2PL: SG + PL PL + PL

3PL: PL + PL

1PLE: SG + PL PL + PL

1PLI: SG + PL PL + SG/PL

2PL: SG + PL PL + SG/PL

3PL: SG + PL PL + PL

gélt̓ex̄d

‘tall’

PL: gigelt̓ex̄d

A

1PL: SG + SG/PL PL + PL

2PL: PL + SG35

3PL: PL + PL

c̓ek̓ʷex̄d

‘short’

PL: c̓ic̓ek̓ʷex̄d

A

1PL: SG + SG/PL PL + PL

2PL: PL + PL

3PL: PL + PL

1PLE: SG + SG/PL

PL + SG/PL

1PLI: SG + SG/PL PL + PL

2PL: SG + PL PL + PL

3PL: SG + PL PL + SG/PL 31 For example, “2PL: PL + SG/PL” would mean that all the combinations with the plural form of the

negative auxiliary verb are accepted in the 2st person plural, “1PL: SG + SG/PL PL + SG/PL” would mean that all the possible combinations are accepted in the 1st person plural regardless of the clusivity, and so on.

32 There are no data of the proximal or absental 3rd person plurals when the semantic head of the predicate is m̓ál̓inix̄ʷ ‘hunter’.

33 All the combinations of 1st and 2nd persons and SG + PL combinations of all the 3rd persons were tested with an alternative plural form hihimas of the semantic head of the predicate.

34 The combination PL + PL in the case of the exclusive 1st person plural is not as certain as the other combinations of the 1st person plural when the semantic head of the predicate is him̓ás ‘chief’.

35 It is unclear whether any other combinations than PL + SG are possible in the 2nd person plural when the semantic head of the predicate is gélt̓ex̄d ‘tall’.

Tero VATTUKUMPU

- 54-

Page 16: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

kʷíxq̓a

‘angry’

PL: kʷíkʷexq̓a

A

1PL: SG + SG/PL

PL + SG/PL

2PL: SG + PL PL + SG/PL

3PL: PL + PL

y̓ák̓iqela

‘sad’

PL: y̓iy̓ak̓iqela

A

1PL: SG + SG/PL

PL + SG/PL

2PL: PL + SG/PL

3PL: PL + PL

hená

‘to sing’

PL: hiʼená

V

1PL: SG + SG/PL

PL + SG/PL

2PL: SG + PL PL + SG/PL

3PL: SG + PL PL + SG/PL

t̓epáƛ

‘to fish’ (FUT)

PL: t̓it̓epáƛ

V

1PL: SG + SG/PL

PL + SG/PL

2PL: SG + PL PL + PL

3PLP: PL + PL

3PLM: PL + PL36

3PLD: SG + PL PL + SG/PL37

3PLA: PL + SG/PL38

máɬela

‘to swim’

PL: mimaɬela

V

1PL: SG + SG/PL PL + SG

2PL: PL + SG

3PL: SG + PL PL + SG/PL

dálanuxʷ

‘to have money’

PL: didálanuxʷ

V

1PL: SG + SG/PL

PL + SG/PL

2PL: SG + PL PL + SG/PL

3PL: PL + SG/PL

1PL: SG + SG/PL PL + PL

2PL: SG + PL PL + PL

3PL: SG + PL PL + SG/PL

36 It is unclear whether any other combinations than PL + PL are possible in the medial 3rd person plural

when the semantic head of the predicate is t̓epáƛ ‘to fish (FUT)’. 37 The combinations SG + PL and PL + SG are not as certain as PL + PL in the case of the distal 3rd

person plural when the semantic head of the predicate is t̓epáƛ ‘to fish (FUT)’. 38 The combination PL + SG is not as certain as PL + PL in the case of the distal 3rd person plural when

the semantic head of the predicate is t̓epáƛ ‘to fish (FUT)’. The consultant also pointed out that the combination PL + SG could perhaps mean that there is one person who is going to fish while saying that ‘They will not fish’. If this was true, it might suggest it to be possible that a combination of the plural form of the negative auxiliary verb and a plural form of the semantic head of the predicate could at least in the absental 3rd person plural have (in some cases) the meaning ‘Not all of them are/do X’. Since the consultant was not sure about the possibility of this kind of interpretation and since there are not any other similar examples, I will not discuss this topic any further in this paper.

Notes on reduplication in the Haisla language ̶ Partial reduplication of the negative auxiliary verb ̶

- 55-

Page 17: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

Some examples of the acceptable and unacceptable combinations listed in Table 5 are

shown below.

(16) a) k̓u=nuxʷ gélt̓ex̄d b) k̓u=nuxʷ gigelt̓ex̄d

AUX.NEG=1EXCL.PL tall AUX.NEG=1EXCL.PL tall.PL

‘We (EXCL) are not tall.’ ‘We (EXCL) are not tall.’

c) *k̓ik̓u=nuxʷ gélt̓ex̄d d) k̓ik̓u=nuxʷ gigelt̓ex̄d

AUX.NEG.PL=1EXCL.PL tall AUX.NEG.PL=1EXCL.PL tall.PL

‘We (EXCL) are not tall.’ ‘We (EXCL) are not tall.’

(17) a) *k̓us=ix kʷíkʷexq̓a b) *k̓ik̓us=ix kʷíxq̓a

AUX.NEG=3SG/PL.PROX angry.PL AUX.NEG.PL=3SG/PL.PROX angry

‘They (here) are not angry.’ ‘They (here) are not angry.’

c) k̓ik̓us=ix kʷíkʷexq̓a

AUX.NEG.PL=3SG/PL.PROX angry.PL

‘They (here) are not angry.’

Possible conclusions that could be made about the data shown in Table 5 are discussed in

the next chapter.

4 Conclusion

As seen from the data shown in the last section of the previous chapter, it seems to be

difficult to make any generalizations of the acceptable combinations of the singular and plural

forms of the negative auxiliary verb and the semantic head of the predicate according to which

any rules for predicting all the acceptable combinations could be established.

At least the following observations could be worth mentioning about Table 5:

1) When there are two sets of acceptable combinations from each consultant, the sets are

different.

2) Consultant B allows different combinations for each person with each stem.

Tero VATTUKUMPU

- 56-

Page 18: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

3) Consultant A seems to treat the noun-like and adjective-like stems similarly in the sense

that (s)he allows only one combination (PL + PL) for both semantic categories in the 3rd

person.

4) It seems to be very uncommon to allow only combinations with a singular form of the

semantic head of the predicate.39

5) The combination of PL + SG seems to be the most unacceptable combination overall.

At the moment, the only generalization that I find possible to make is number 3) above.

This is very interesting because if this generalization could be supported by more data from

the consultant A, it could potentially provide valuable evidence for the debate over the topic of

lexical categories in Haisla, i.e. the possible division of the roots into nouns, adjectives and

verbs. One thing making this generalization less meaningful is the fact that there was a great

degree of uncertainty about the acceptability of the combinations among the consultants

during the elicitation before the final set of acceptable combinations was formed. Also, the

number of collected examples is perhaps still too low to come to any conclusion.

On the other hand, there were also comments from the consultants pointing out that some of

the acceptable combinations must be understood to be acceptable only in certain contexts.

Because of this, it should be understood that “acceptability” of a given combination in this

paper means that the combination can be used but not always. This of course raises a question

about the actual meaning of the collected data and the adequacy of the methods used to collect

it. Namely, if the data had been collected by taking the context more into consideration, it

might be much more consistent, if we assume that there were underlying rules governing what

kind of combinations were accepted with certain stems.

I will conclude this paper by remarking that for now there is not enough evidence to make

any generalizations about the acceptability of the combinations of the singular and plural

forms of the negative auxiliary verb and the semantic head of the predicate, and that

combinations seem to be more or less in a free variation in the sense that there is no common

tendency between the consultants, stems, nor persons. This may be a result of a diachronic

change in the language as one of my consultants pointed out to me that the plural form of the

negative auxiliary verb is used more often by the oldest generation of the native speakers40, 41.

39 This could of course be due the fact that only one kind of combination with a singular form of the

semantic head of the predicate (PL + SG) can potentially be accepted in the 2nd and 3rd persons since a combination of SG + SG would lead to an interpretation of a singular subject, as pointed out in the Chapter 2.

40 The consultant referred to the way of speaking of the oldest generation as “the old Haisla”.

Notes on reduplication in the Haisla language ̶ Partial reduplication of the negative auxiliary verb ̶

- 57-

Page 19: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

If it was supported by more evidence, it might be possible to assume that the usage of the

plural form was much more systematic in the past, which could imply that the system that

might have been consistent earlier has now started to decline. Collecting data about

combinations of singular and plural forms of the semantic head of the predicate and other

auxiliary verbs could possibly shed some light to the big picture of the reduplication and plural

forms of the auxiliary verbs.

List of abbreviations and symbols

- Affix boundary DIS / D Distal

= Clitic boundary EXCL / E Exclusive

∅ Zero morph FUT Future tense

Possible combination INCL / I Inclusive

Impossible combination LS Lexical suffix

1 First person MED / M Medial

2 Second person N Noun-like stem

3 Third person NEG Negative

A Adjective-like stem POSS Possessive suffix

A Absental PROX / P Proximal

AUX Auxiliary verb PL / PL Plural

CAP Capability SG / SG Singular

DEIX Deixis suffix V Verb-like stem

DEM Demonstrative pronoun

41 Both my consultants are a little bit over or under 70 years old.

Tero VATTUKUMPU

- 58-

Page 20: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

References

Bach, Emmon (1990a) Stem Extensions in Haisla. International Conference on Salishan and

Neighboring Languages 25.1-16.

Bach, Emmon (1990b) A Haisla book. Unpublished draft.

Bach, Emmon (1995) A note on quantification and blankets in Haisla. In: Emmon Bach, Eloise

Jelinek, Angelika Kratzer and Barbara H. Partee (eds.) Quantification in Natural Languages

13-20. Dodrecht: Kluwer.

Bach, Emmon (2001) Building words in Haisla. University of Massachusetts Occasional

Publications 20: 51-73.

Bach, Emmon (2006) Paradigm Regained: Deixis in Northern Wakashan. SOAS Working

Papers in Linguistics 14. 267-281.

Bach, Emmon and Bates, Reed (1971) Some notes on Xaʼisla. In: James E. Hoard and Thomas

M. Hess (eds.) Studies in Northwest Indian Languages 1-11. Sacramento: Sacramento

Anthropological Society.

Bach, Emmon; Robinson, Dora and Robinson, Rose (n.d.) Wisenis X̄aʼislak̓ala. Unpublished

learning material.

Lincoln, Neville J. and Rath, John C. (1986a) Phonology, Dictionary and Listing of Roots and

Lexical Derivates of the Haisla Language of Kitlope and Kitimaat [sic], B.C. Volume 1.

Canadian Ethnology Service Paper 103. Ottawa, ON: National Museum of Canada.

Lincoln, Neville J. and Rath, John C. (1986b) Phonology, Dictionary and Listing of Roots and

Lexical Derivates of the Haisla Language of Kitlope and Kitimaat [sic], B.C. Volume 2.

Canadian Ethnology Service Paper 103. Ottawa, ON: National Museum of Canada.

Vink, Hein (1977) A Haisla phonology. International Conference on Salish Languages 12:

111-131.

Notes on reduplication in the Haisla language ̶ Partial reduplication of the negative auxiliary verb ̶

- 59-

Page 21: Title Notes on reduplication in the Haisla …...Title Notes on reduplication in the Haisla language --Partial reduplication of the negative auxiliary verb--Author(s)

Tero VATTUKUMPU

ハイスラ語における重複について一否定助動詞における部分重複 -

ワットワクンフ。

テロ

要旨

本稿では、 カナダで話されている先住民言語であるハ イスラ語の否定助動詞には、

先行研究における記述に反して、部分重複によって形成される複数形があることを、

筆者が集めたデー タに基づいて示す。 ハ イスラ語において、 主語の複数性は述語で

複数形を使うことによって表せる。 述語が助動詞も含めている場合に、 本動詞と助

動調のどちらかまたはその両方を複数形にすることができる。 いくつかの述語の場

合に否定助動詞と本動詞の単数形と複数形がどの組み合わせで出現できるかを調査

した。 集めたデー タによれば、 その組み合わせに関しては、 一般化ができないよう

である。

キ ー ワ ー ド: ハ イスラ語、 部分重複、 否定助動詞、 複数、 語根拡張

- 60 -